Richard Hanania recently highlighted an NPR/Marist poll showing Republicans slightly more likely than Democrats to believe that “Americans may have to resort to political violence.” His observation was correct — not because it revealed anything about partisanship, but because it exposed the question itself as the real instrument of bias. The subtle ambiguities in phrasing aren’t incidental; they are the entire point.
There is almost never a good reason to ask a question this way: “Americans may have to resort to political violence to get the country back on track.” It’s not a probe of moral conviction. It’s a psychological Rorschach — an engineered ambiguity that measures framing effects, not beliefs.
Look carefully at what the pollsters did here. The question uses three linguistic tricks to inflate agreement while preserving plausible deniability:
Diffused Agency.
The word “Americans” is an absolution. It’s not you, it’s someone else like you — a member of the moral collective. It turns personal responsibility into abstract observation: “Some people might have to do what’s necessary.” That phrasing grants moral distance to the respondent while still activating identification with the group. It’s permissioned empathy for hypothetical violence.Implied Necessity.
The phrase “may have to” presupposes legitimacy. It encodes inevitability, as though political violence could be a reluctant duty forced upon good citizens. The respondent is not endorsing violence — merely acknowledging its tragic possibility. In reality, this is assent laundering.Normative Framing.
The clause “to get the country back on track” embeds a presupposition: the country is off track. The question begins with a grievance baked in. Anyone dissatisfied with the state of the nation — which is most people — is subtly guided toward agreement, regardless of their actual view on violence.
Put together, these linguistic sleights transform a moral question (Is violence justified?) into a sentimental one (Are you frustrated enough to believe someone might have to act?). The difference is everything.
When you replace “Americans” with “people like me,” or worse, “members of my political party,” support collapses. The abstraction of responsibility is what keeps the poll results inflated. This isn’t measuring extremism; it’s measuring distance from culpability.
Pollsters know this. They rely on subtle manipulations of agency and necessity to manufacture headlines: “Republicans more likely to endorse political violence.” The intent isn’t to reveal; it’s to polarize. The ambiguity isn’t a flaw — it’s the product.
A neutral formulation would read:
“Do you personally believe that the use of physical force for political goals can ever be justified in the United States today?”
Ask that, and you’ll find near-universal rejection across parties — a boring, stable result unfit for the outrage economy. The sensational asymmetry exists only inside the linguistic fog.
Pollsters don’t measure beliefs. They design them. And when you control the phrasing of the question, you control the moral topography of the answer.